Applied Biostatistics: Comprehensive Hypothesis Testing Analysis

This report provides a comprehensive overview of hypothesis testing, a crucial method for evaluating population data hypotheses using statistical methods. It begins by defining hypothesis testing and explaining its importance in determining sample characteristics, establishing population parameters, and comparing data consistency. The report then delves into the five key steps of hypothesis testing: defining null and alternative hypotheses, setting the significance level, calculating test statistics and p-values, and drawing conclusions. It clarifies the differences between one-tailed and two-tailed hypotheses, and the implications of Type I and Type II errors. Furthermore, the report discusses data preparation techniques, including data collection methods, editing, coding, and exploratory data analysis. The report concludes by emphasizing the importance of hypothesis testing in research and its role in making informed decisions about research findings.

Running head: HYPOTHESIS TESTING 1
Applied Biostatistics: Hypothesis Testing
Student Name
Institution Affiliation

HYPOTHESIS TESTING 2
Hypothesis Testing
Introduction
Hypothesis testing is a method of evaluating a hypothesis about populated data using
statistical methods. The reasons for conducting a population hypothesis include determining a
characteristic of a sample, obtaining a sample randomly from a population, establishing a
population parameter and conducting a comparison of obtained data to predict its
consistency. Before testing a hypothesis it is essential to understand its meaning. A
hypothesis is a tentative verification of a solution to a problem. The rationale is based on
empirical verification of two or more variables in a specified frame of reference and it does
not have to be correct. While it is a prediction about what is expected in a study, it involves
the population, the variables and the variables relationship. The goal of a research hypothesis
is to determine if it is right or wrong by weighing various factors and settling on one that can
result in the desired outcome.
Hypothesis Testing’s Five Steps
The two types of hypothesis are; one-tailed hypothesis and two-tailed hypothesis.
There must be accountability for the variables and the sample size in order to deduce whether
the expected changes are meaningful. Both the null and the alternative hypotheses must be
specified before the significance level is set, the test statistic and corresponding P-Value have
to be calculated before lastly drawing a conclusion. The null hypothesis (H0) exhibits no
change in the general population and thus the independent variable does not affect the
dependent variable (Neuendorf, 2016). This hypothesis is a statement that stands to be
disproved. Examples of these variables include the fact that all dogs have four legs; as well
no relationship exists between a type of injury and if the patient was given an IV injection
before being hospitalized.

HYPOTHESIS TESTING 3
The alternative hypothesis (H1) attests that there are changes, differences and
relationships in the general population where the independent variables have an effect on the
dependent variables. Researchers are more interested in proving this hypothesis and is
normally either one sided or two sided. Two sided tests are a common feature because they
provide more evidence against the null hypothesis (Neuendorf, 2016). Some examples of this
type of hypothesis include; an intubation success rate is different in patients under treatment,
the association between a type of injury and if a patient received an IV before hospitalization
and lastly the time taken by the doctors for resuscitation from a cardiac arrest is lower than its
control.
Setting the significance level literally gives a minimal chance of acceptance for the
other hypothesis while the null hypothesis remains positive. When the level of significance is
small, the verification required to decline the null hypothesis increases. Calculating the test
value and the corresponding P- value is the next step and a test statistic is meant to examine
the association between variables using a confidence interval while the p-value describes the
likelihood of getting a sample statistic and is decided through the test statistic outcome
(Neuendorf, 2016). The conclusions on the hypothesis result from the p-value and the level of
significance. For example if the p-value is 0.01, it will happen only once in 100 times if the
null hypothesis is positive implying that it is unlikely to happen. Likewise a p-value of 0.75
occurs 75 out of 100 times by chance when the null hypothesis is positive. However, it is
necessary to note that your sample size directly affects the p-value and it is important to plan
your sample size ahead of time as well as verify the relevance of the results.
The conclusion aims at translating an alternative hypothesis into a statement as either
the outcome is statistically significant or not to enable deductions as to whether the null
hypothesis is to be accepted or rejected.

HYPOTHESIS TESTING 4
Define Null and Research Hypotheses.
A null hypothesis is a declaration of no variation given at the beginning of a scientific
investigation that helps to simplify an approach by conforming to the concept of falsification
(Ary et al., 2018) This type of hypothesis does two things; first it constructs a question using
a falsification outlook and secondly it tries to interpret data patterns in a simplified way.
Furthermore, it is paired with alternatives such that the null implies that there is no difference
among variables, while the alternative states otherwise. In the event that the alternative is the
case, a query of how these patterns relate to the scientific hypothesis under test is raised.
A research hypothesis describes the intended outcome or an informed guess of a
researcher. The researcher believes that this hypothesis is supported by available data. An
example of a research hypothesis is that private high school graduates’ fathers are of high
status occupations than public school graduates.
Preparing Data for Hypothesis Testing.
Data used for hypothesis testing is surveyed and collected using available data
collection methods which may include questionnaires, observation and interviews before
subjecting it to pilot testing (Roiger, 2017). Once it is collected, editing and coding
procedures are used to ascertain its integrity by doing normalcy and outlier checks. Data
collection begins by selection of field personnel, training them, supervision and eventually
evaluation. The researcher must have clarity of the study to be undertaken. The nature and
scope of the study follows a procedure that involves data editing, coding, entry as well as data
cleaning to make it ready for eventual data analysis (Chambers, 2017).
Exploratory Data Analysis as a Prelude to Hypothesis Testing.

HYPOTHESIS TESTING 5
Exploratory data analysis involves the use predetermined descriptive procedures that
describe the structure of data. Besides using these methods, it also uses exploits the intuition
of the research expert in investigating the data structure. Contrary to the confirmatory mode
which relies on classical analysis to confirm and reformulate theory using parameter
estimation and hypothesis testing, it harnesses informal data examination thereby being
considered as a prelude to both classical statistical analysis and hypothesis testing (Palms et
al., 2018). It is supplementary to the confirmatory mode and is useful in analyzing economic
phenomena.
Type I and II Errors.
When type I errors occur, a null hypothesis is positive but it is not accepted. It is also
called an error of the first kind. This error gives a false assertion by giving an indication that
something is present when it is actually absent and is used to test single conditions. A type II
error on the other hand is second kind error which happens when there is a false null
hypothesis but fails to be rejected by mistake. It misses to confirm what is present (Urbano,
Lima & Hanjalic, 2019).

HYPOTHESIS TESTING 6
References
Ary, D., Jacobs, L. C., Irvine, C. K. S., & Walker, D. (2018). Introduction to research in
education. Cengage Learning.
Chambers, J. M. (2017). Graphical Methods for Data Analysis: 0. Chapman and Hall/CRC.
Palm, R., Sorg, C. G., Ströbel, A., Gerritsen, D. L., & Holle, B. (2018). Severe Agitation in
Dementia: An Explorative Secondary Data Analysis on the Prevalence and Associated
Factors in Nursing Home Residents. Journal of Alzheimer's Disease, (Preprint), 1-8.
Neuendorf, K. A. (2016). The content analysis guidebook. Sage.
Roiger, R. J. (2017). Data mining: a tutorial-based primer. Chapman and Hall/CRC.
Urbano, J., Lima, H., & Hanjalic, A. (2019). Statistical Significance Testing in Information
Retrieval: An Empirical Analysis of Type I, Type II and Type III Errors. arXiv
preprint arXiv:1905.11096.